What we do
To be the best at what we do, our core technology uses consensus algorithmic profiling to identify and rank B cell epitopes at the surface of a target protein. Obtaining maximum value from our analysis of proteins of vastly differing structure and function requires human intervention for results evaluation in the context of acquiring new molecular immunoinformatic knowledge for use in protein annotation. Crucial to optimising a workflow for predicting antibody epitopes is cross-referencing to critically evaluated information available for the target protein of interest, which also assists tailoring our services to individual target proteins. Thus, our epitope discovery engine combines using a range of sequence analysis programs with proprietary empirical analysis of protein sequence and structure, to provide epitope prediction and relevant epitope contextual analysis and annotation of target proteins.
Consensus antibody epitope prediction
Epitope Informatics' key proprietary technology is a consensus epitope prediction platform used for the discovery and characterisation of antibody epitopes. Our bioinformatic analysis typically proceeds as a sequence of predetermined data processing tasks, generating a list of consensus predicted epitopes ranked in order of their likely antigenicity.
Algorithms used for target sequence analysis include:
- Antigenicity (3 algorithms).
- Flexibility (backbone chain; 1 algorithm).
- Hydrophilicity (2 algorithms).
- Hydrophobicity (5 algorithms).
Further target sequence analysis and annotation, which in part is guided by our observation that B cell epitopes are often found in association with protein sequence motifs and can point to epitope targets of protein function, includes:
- The effect of varying algorithm window size on target protein profiles.
- Signal/cell sorting sequences.
- N- and C-terminus regions.
- Secondary structure.
- β-turns and disulphide bonds.
- Protein domains.
- Subcellular location.
- Transmembrane spanning and other hydrophobic regions.
- Low complexity protein sequence.
- Internal sequence repeats.
- Intrinsically disordered protein structure.
- Protein aggregation potential.
- Protein degradation motifs.
- PROSITE pattern motifs.
- SUMOylation motifs.
- Other functional motifs (of viruses, bacteria, fungi, eukaryotes ...).
- Miscellaneous target protein analysis.
- Critical evaluation of information available for a target protein.
- Specialist databases for information relating to target protein and predicted epitopes.
Consensus predicted epitopes show:
- High values for predicted antigenicity, flexibility and surface location.
- A high degree of association with predicted β-turns (secondary structures frequently associated with antigenic sites).
- Low hydrophobicity and high hydrophilicity values.
Hydrophobic residues prefer the buried location, but about 30% are found at the protein surface in association with structures including epitopes and hydrophobic pockets and clefts, and are needed for functions such as protein-protein interaction and ligand binding. That not all antigenic sites are hydrophilic underscores the value to epitope prediction of our consensus approach with expanded target sequence analysis and annotation.
Epitope mapping to target protein 3-D structure
Our protein structure analysis and modelling platform includes use of high performance graphics workstations equipped with stereoscopic 3-D viewing hardware. Reducing 3-D molecular model viewing to two dimensions is a form of information reduction and downgrade. Thus, visualisation and the reporting of predicted epitopes mapped to protein 3-D structures is greatly enhanced by our use of stereoscopic 3-D viewing hardware, especially for conformational, cryptic and quaternary epitopes. We use molecular viewing and annotation software that includes Discovery Studio and other software used under license from BIOVIA (formerly Accelrys and Oxford Molecular). This platform allows our provision of an enhanced protein structure analysis and modelling service that is optimised for antibody epitope prediction, epitope contextual analysis and target protein annotation.
Mapping and contextual analysis of predicted antigenic regions on protein 3-D structure:
- Increases the accuracy of epitope prediction.
- Confirms surface location of a predicted epitope and establishes the degree of epitope surface exposure.
- Determines epitope structure and boundaries.
- Aids the identification of potential conformational/discontinuous epitopes.
In our lengthy experience, linear epitopes may come together at the surface of a protein to form conformational epitopes. The extent to which this occurs across a proteome is currently unknown, but it is a subject of investigation at Epitope Informatics.
For up to six consensus predicted epitopes per target protein, our analysis includes:
- Degree of epitope surface location and residue side-chain exposure.
- Degree of epitope flexibility.
- Epitope association with β-turns (secondary structures frequently associated with antigenic sites, particularly when they are adjacent to β-sheets or α-helical structures in regions of hydrophilicity and polypeptide chain flexibility) and other structures associated with antigenic sites.
- Conformational epitopes.
- Rank of epitopes as likely antigens.
- Epitopes mapped to multiply aligned target protein and homologue sequences and structures.
- Protein database searching optimised for use with predicted epitopes as search query strings.
One of our goals is to remain relevant to scientists' needs by responding constructively to feedback. Therefore, we would appreciate any input you are willing to share regarding services we might provide in the future.
Possible future service options include:
- Epitope conservation analysis performed in a high performance cloud computing environment.
- Modelling target protein epitope mutation and mutation propensity.
- Epitope remodelling.
- Target protein and antibody interaction analysis (docking/molecular dynamics).
- Target protein antibody engineering.
- T cell epitope prediction and analysis.